Bonjour

With the growth of 24h distribution application developed by Chappiebot Inc., we are in need of an algorithm to test many models with the same purpose to see whether which one is the best. The traditional method is A/B Test. However, it is observed that this old-fashioned way becomes inefficient... [Read More]

In the process of building news distribution platform for 24h News, we realize the importance of a Named-entity Recognition (NER) Model. It helps us to identify the user’s preferences, then we could amplify the effectiveness of the platform. In this blog, we will explain our approach in more details. [Read More]

Nowadays, deployment plays a major role in applying Deep Learning in daily life. We could build an algorithm and develop a demo in our development machine. However, to put these algorithm in production at large scale, using Deep Learning framework offered by large corporations is a requirement. Today, I will... [Read More]

In OtoNhanh.vn, there is a section which allows the users to compare one model versus another in the aspect of specifications and the exterior as well as interior views. Normally, we will grab randomly 2 images of 2 models from the same category. Nevertheless, we observe that this approach doesn’t... [Read More]

In the last blog about Distributed TensorFlow, we have provided some fundamental knowledge of Distributed Training in this framework. However, it is not enough if we want to apply it efficiently. Today, we will provide some additional tricks to make use of Distributed Computing better in OtoNhanh.vn. [Read More]